Edward O'Callaghan has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/33931 )
Change subject: flashchips.c: Sort file by vendor and model ......................................................................
Patch Set 3:
Patch Set 3: Code-Review-1
(1 comment)
This doesn't seem to adhere to the currently defined order. I wouldn't mind to change that definition, though. Also, the post-merge plans are better discussed on the mailing list. There are other people that have plans too :) For instance during the OSF Hackathon in June [1], we discussed making it an easier to handle ASCII file, that can (optionally) be compiled into C.
[1] http://docs.google.com/document/d/18qKvEbfPszjsJJGJhwi8kRVDUG3GZkADzQSH6WFsK...
Hi Nico,
Long time, hope you have been well?
I actually had a similar idea of a mini DSL that describes chips and breaking up the configs into a per-vendor thing. It would be good to discuss that more as I should be back around more again these days!
Alan worked on this in collaboration with me. I am trying to get rid of the Chromium flashrom fork and get all the extra chip (+chipset) support from that into upstream. Unfortunately they have diverged *significantly* [9+ years] so it has been a real battle. Alan was able to contrive a tool that allowed us to reconcile many of the differences in this specific file however there are some conditions to be able to reach the point of convergence, this obviously being one of them.
More precisely, I think if we can come up with a ordering function that produces reproducible results of ordering that would also work well however this was Alan's best first go at getting a ordering that works enough to make progress merging the two flashchips.c implementations.
Once we have all the chip definitions in one place again we can move forward with greater and better plans on a nicer representation. I hadn't seen that doc until now however I don't see a problem with this specific blocking that plan from happening, in fact this could help aid in that since the ordering and cleanups make the file consistent enough now to be parsed up into a data structure and written out to ASCII or whatever representation we decide to choice.
Cheers, Edward.